Topological photonic lattices with long-range interaction break the channel bandwidth limit

Topological physics deals with physical quantities that remain unchanged under deformation. Due to its potential for noise-insensitive calculations and signal processing, it has attracted great interest in various fields of physics – photonics, quantum computing, solid-state physics, acoustics and electronic circuits.

In the field of photonic integrated circuits, whose goal is to perform computations by adjusting light fluxes, there has been concentrated effort over the last decade to implement noise-resistant signal processing using topologically protected light states.

One of the biggest challenges in the practical implementation of this research direction is overcoming the bandwidth limitations of signal processing. When implementing topological properties using neighboring interactions between optical elements of a two-dimensional system, there is a trade-off relationship between the number of signal channels and the channel bandwidth determined by the grating band gap. This trade-off prevents an increase in the overall information capacity (= channels × bandwidth).

In a new article published in Light: Science and ApplicationsA team of scientists led by Professor Sunkyu Yu and Professor Namkyoo Park from the Department of Electrical and Computer Engineering, Seoul National University (Korea) and colleagues have implemented a defect-robust multi-channel signal processor by tuning long-range interactions in topological photonic lattices.

Unlike conventional topological systems that rely solely on nearest-neighbor interactions, the researchers’ system uses hardware that allows significant interactions over long distances, enabling effective overlap of conventional lattice structures within the two-dimensional plane.

This “lattice overlap” technique allows the tuning of topological invariants – here “the Chern numbers” – while preserving the band gap of an original lattice. This tunable Chern number provides multi-channel topologically protected edge modes and finally breaks the trade-off relationship between the number of signal channels and the channel bandwidth.

In the paper, the scientists implemented their lattice overlap strategy in the famous Hofstadter model using system parameters available in conventional silicon photonics, which they analyzed using the Tidy3D software.

They demonstrated the incoherent optical functionality – a multi-channel wave splitter for light with random phases and amplitudes – that is extremely robust against various types of disturbances. The result shows that their approach enables noise-insensitive signal processing with improved information capacity.

“To date, topological photonic circuits have typically been implemented under the condition of nearest-neighbor interactions between optical elements. Although some efforts have been made to analyze the effects of long-range interactions, due to the nature of electromagnetic waves, these interactions were assumed to be much weaker than the interactions between neighboring elements.

“In our work, we develop integrated photonic platforms that can achieve stronger interactions over long distances than neighboring ones. This approach enables the implementation of the ‘lattice overlap’ design, effectively realizing the multiple overlap of the conventional lattices within a two-dimensional plane.

“This overlapping lattice provides the design freedom to achieve arbitrary Chern numbers while maintaining the band gap width. We demonstrated the result using the Hofstadter model as an example. The obtained Chern number manipulation enables multi-channel, topologically protected edge modes in the wide band gap, thus solving the channel bandwidth problem.

“Our result therefore enables robust and powerful signal processing in photonic integrated circuits, which are widely used in AI accelerators and quantum computing.

“Long-distance interactions partially emulate higher-dimensional physics, as shown by the increased coupling levels at each node. Therefore, we fundamentally believe that the most significant impact of our research would be the effective modeling of higher-dimensional physics on a two-dimensional level. This breakthrough could pave the way for the implementation of topological phenomena in complex networks, which is one of the ultimate goals of our research,” the scientists explain.